images CHAPTER 8

Optimization of Time Series Using Parallel, Adaptive, and Neural Techniques

J. A. GÓMEZ, M. D. JARAIZ, M. A. VEGA, and J. M. SÁNCHEZ

Universidad de Extremadura, Spain

8.1 INTRODUCTION

In many science and engineering fields it is necessary to dispose of mathematical models to study the behavior of phenomena and systems whose mathematical description is not available a priori. One interesting type of such systems is the time series (TS). Time series are used to describe behavior in many fields: astrophysics, meteorology, economy, and so on. Although the behavior of any of these processes may be due to the influence of several causes, in many cases the lack of knowledge of all the circumstances makes it necessary to study the process considering only the TS evolution that represents it. Therefore, numerous methods of TS analysis and mathematical modeling have been developed. Thus, when dealing with TS, all that is available is a signal under observation, and the physical structure of the process is not known. This led us to employ planning system identification (SI) techniques to obtain the TS model. With this model a prediction can be made, but taking into account that the model precision depends on the values assigned to certain parameters.

The chapter is structured as follows. In Section 8.2 we present the necessary background on TS and SI. The problem is described in ...

Get Optimization Techniques for Solving Complex Problems now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.